[time-nuts] Re: Is this the right way to compare the short term accuracy of two frequency counters?

[Poul-Henning Kamp]

--------
Tom Van Baak writes:

> To determine the noise floor of a counter it's best to use different 
> sources, independent in frequency and also independent in phase. That 
> rules out dividers or multipliers. The phase needs to "sweep" across a 
> wide range.

A neat way to do this:


    Clock --+-- (ext timebase) HP3336 ----->
            |                                  TI-counter
            +------------------------------>

Set the HP3336 to the same frequency as the clock source
and use it to vary the relative phase of the two signals
into the counter.

The main advantage is that you can make many measurements at each
of the 360 phase differences and thus get both avg+stddev along
the curve.

Any unlinearities in the HP3336 obviously mix into the resulting
measurements, but as far as I can tell, it performs much better
than any of my counters.

That is probably not unrelated to it being built to measure "time
distortion" on telephony channels frequency-stacked on coax cables :-)

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.